I am about to use ELISA tests to measure pro-inflammatory molecules from human blood, taken from dementia patients. Is it best to use serum or plasma for this tests?
Form the viewpoint of lab chemist, plasma is always the best choice. Why?
In serum the coagulation cascade was activated. There are many proteases included, proteases which are not always highly specific. They can cleave your analyte as well.
If you would like to earn serum you have to wait some time until the coagulation is finished. Why to wait for such a long time when plasma can be earned without any delay? Most cytokines have short half life times due to the proteolytic digestion.
If the fibrinogens are a problem for you, use serum. If not, use plasma. But if as you said, you are working with pro-inflamatory molecules, think that they may to increase fibrinogen levels in your samples.
Serum or plasma is fine. We used for TNF, IL-6 etc all the time. I would use whichever is easier to get. Typically there is a larger serum volume than plasma.
If it is possible, I would suggest that you see what the recovery rate is like for a spiked sample of your pro-inflammatory molecule in both serum and plasma, then at least you'll know whether it is the ELISA compatibility or the pre-processing which might be affecting your results (if indeed there is any discrepancy between serum and plasma readings).
I would use the serum provided of course that you are sure that the molecule measured is not consumed during coagulation. You avoid the introduction of an anticoagulant that we can never be sure it does not interfere.
I think the choice should have been made at the moment you designed the research as it influences the way samples are collected. At that time both choices could be right; when I don't have any rationale to choose between plasma or serum, I search literature about methodology used by other researchers measuring stuff I want to evaluate. I totally agree with Nicole about the idea of using serum NOW because you should use an anticoaugulant when collecting blood and a protease inhibitor (such aprotinin) if you plan to measure proteins.
Many thanks for all of your comments. I think it is interesting that this is a point of debate even though these tests are used so commonly. I have planned to collect both plasma and serum so I may be able to do both and investigate any difference.
Form the viewpoint of lab chemist, plasma is always the best choice. Why?
In serum the coagulation cascade was activated. There are many proteases included, proteases which are not always highly specific. They can cleave your analyte as well.
If you would like to earn serum you have to wait some time until the coagulation is finished. Why to wait for such a long time when plasma can be earned without any delay? Most cytokines have short half life times due to the proteolytic digestion.
I fully agree with Stephan that if you want to study inflammatory molecules circulating in blood, plasma is the best material to use for this type of assays. Indeed, when you prepare serum, plasma is activated by a cascade of reactions that lead to the conversion of fibrinogen into fibrin as a result of the generation of thrombin, and other proteases. In addition, thrombin will activate the platelets that will release in serum all their content (including inflammatory and anti-inflammatory mediators), screwing up your data (artefact). The blood should therefore be collected with an anticoagulant (citrate based solution, EDTA, or heparin), kept at 22°C +/- 2°C to avoid platelet activation by cold temperature; and centrifuged within about 6 hrs to isolate the plasma from cellular elements. Plasma, free of platelets, can then be frozen.